SpikeGPU  1.0.0
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros
spike::SpmvCusp< Matrix > Class Template Reference

Default SPMV functor class. More...

#include <spmv.h>

Public Member Functions

 SpmvCusp (Matrix &A)
 
double getTime () const
 Cummulative time for all SPMV calls (ms). More...
 
double getCount () const
 Total number of calls to the SPMV functor. More...
 
double getGFlops () const
 Average GFLOP/s over all SPMV calls. More...
 
template<typename Array >
void operator() (const Array &v, Array &Av)
 Implementation of the SPMV functor using cusp::multiply(). More...
 

Detailed Description

template<typename Matrix>
class spike::SpmvCusp< Matrix >

Default SPMV functor class.

This class implements the default SPMV functor for sparse matrix-vector product, using the cusp::multiply algorithm.

Template Parameters
Matrixis the type of the sparse matrix.

Constructor & Destructor Documentation

template<typename Matrix >
spike::SpmvCusp< Matrix >::SpmvCusp ( Matrix &  A)
inline

Member Function Documentation

template<typename Matrix >
double spike::SpmvCusp< Matrix >::getCount ( ) const
inline

Total number of calls to the SPMV functor.

template<typename Matrix >
double spike::SpmvCusp< Matrix >::getGFlops ( ) const
inline

Average GFLOP/s over all SPMV calls.

template<typename Matrix >
double spike::SpmvCusp< Matrix >::getTime ( ) const
inline

Cummulative time for all SPMV calls (ms).

template<typename Matrix >
template<typename Array >
void spike::SpmvCusp< Matrix >::operator() ( const Array &  v,
Array &  Av 
)
inline

Implementation of the SPMV functor using cusp::multiply().


The documentation for this class was generated from the following file: